In the data below the ID's is sorted in Waves (years). The Retirement variable can take 3 values, 0,1 and 2. The Weekly_Hours variable can take any value. A respondent is NOT retired if he/she has answered 0 or if he/she has answered 2 along with Weekly_Hours>=20. The opposite means retired (that is Retirement=1 or Retirement=2 & Weekly_Hours<20). I want to drop all id's for a respondent if he went on retirement and back to work in the wavespan he was in the data.
So if a respondent over the course of the waves goes from:
- Retirement = 0 to Retirement = 1 to Retirement = 0
- Retirement = 0 to Retirement = 1 to Retirement = 2 & Weekly_Hours>=20
- Retirement = 0 to Retirement = 2 & Weekly_Hours<20 to Retirement = 0
- Retirement = 0 to Retirement = 2 & Weekly_Hours<20 to Retirement = 2 & Weekly_Hours>=20
- Retirement = 2 & Weekly_Hours>=20 to Retirement = 1 to Retirement = 0
- Retirement = 2 & Weekly_Hours>=20 to Retirement = 1 to Retirement = 2 & Weekly_Hours>=20
- Retirement = 2 & Weekly_Hours>=20 to Retirement = 2 & Weekly_Hours<20 to Retirement = 0
- Retirement = 2 & Weekly_Hours>=20 to Retirement = 2 & Weekly_Hours<20 to Retirement = 2 & Weekly_Hours>=2
ID where the above happens I want to drop the hold group with the same ID. Hope it make sense if not then please ask for more information.
Code:
* Example generated by -dataex-. To install: ssc install dataex
clear
input double(HHIDPN Waves Retirement Weekly_Hours)
1010 1 0 40
3010 1 0 40
3010 2 0 40
3010 3 2 32
3010 4 2 30
3010 5 2 24
3010 6 2 32
3010 7 2 32
3010 8 1 0
3010 9 1 0
3010 10 1 0
10013010 1 0 45
10013010 3 0 40
10013010 4 0 20
10013010 5 1 0
10013010 6 1 0
10013010 7 1 0
10013010 8 1 0
10013010 9 1 0
10013010 10 1 0
10038010 1 0 55
10038010 2 0 50
10038010 3 0 50
10038010 4 0 40
10038010 5 0 40
10038010 6 2 30
10038010 7 0 30
10038010 8 2 20
10038010 9 2 20
10038010 10 2 10
10050010 1 0 35
10050010 2 0 35
10050010 3 0 35
10050010 4 0 35
10050010 5 0 35
10050010 6 0 35
10050010 7 0 35
10050010 8 0 35
10050010 9 0 35
10050010 10 0 35
10063010 1 0 0
10063010 3 0 20
10063010 7 1 0
10063010 8 1 0
10063010 10 1 0
10094010 1 0 30
10097010 1 0 45
10097010 2 1 0
10109020 1 0 30
10109030 1 0 72
10109030 3 0 48
10109030 4 2 20
10109030 5 2 15
10109030 6 2 20
10109030 7 2 20
10109030 8 2 20
10109030 9 2 20
10109030 10 2 30
10114010 1 0 35
10114010 2 1 0
10114010 3 1 0
10114010 5 1 0
10155010 1 0 40
10155010 2 0 40
10155010 3 0 40
10155010 4 0 40
10210010 1 0 30
10210010 2 0 0
10210010 3 0 40
10210010 4 0 20
10210010 5 0 20
10210010 8 1 0
10210020 1 0 0
10210020 2 1 0
10210020 3 0 0
10210020 4 0 40
10210020 6 0 40
10210020 7 2 0
10210020 8 1 0
10210020 9 1 0
10210020 10 1 0
10225010 1 0 37
10225010 2 0 0
10225010 3 1 0
10225010 5 1 0
10225010 6 2 0
10225010 7 2 0
10225010 8 1 0
10225010 9 1 0
10225010 10 1 0
10237020 1 0 40
10237030 1 0 35
10240010 1 0 40
10240010 2 2 0
10240010 6 1 0
10299010 1 0 40
10299010 2 0 40
10299010 4 0 55
10299010 5 0 45
10299010 6 2 10
end
0 Response to Dropping every ID when respondent go from "work" to "retirement"
Post a Comment